R/EssentialJSONSerializer.R
In VermeirJellen/essentialjsonserializer_r: JSON (de)serialization of complex R-objects
Sys.setenv(tz="UTC")
#' JSON serialization of complex R-objects \cr
#'
#' The intention of the package fuctionality is to (de)serialize language specific complex objects towards generic
#' JSON-datatype objects. The JSON conversion allows for flexible cross language communication when the package is implemented
#' in different languages. This package provides an R implementation of the (de)serialization functionality.
#'
#'
#' Currently, 6 generic complex JSON-datatypes are provided. Corresponding R-mappings are defined as follows:
#'
#' \itemize{
#' \item JSON datetime - maps to POSIXct objects
#' \item JSON timeseries - maps to 1-dimensional xts objects
#' \item JSON timeseries_n_dim - maps to n-dimensional xts objects (n > 1)
#' \item JSON series: - maps to 1-dimensional dataframe objects
#' \item JSON dataframe: - maps to n-dimensional dataframe objects (n > 1)
#' \item JSON matrix: - maps to matrix objects
#' }
#'
#' The package provides the user with the EssentialJSONSerializer reference class. Following functions
#' are provided:
#' \itemize{
#' \item RToList: Convert complex R structures to a generic datatype list structure.
#' \item RToJSON: Convert complex R structures to a generic datatype JSON structure.
#' \item RFromList: Convert generic datatype list structure to complex R-specific objects.
#' \item RFromJSON: Convert generic datatype JSON structure to complex R-specific objects.
#' }
#'
#' Examples: \cr
#' - https://github.com/VermeirJellen/essentialjsonserializer_r \cr
#' - http://EssentialDataScience.com/blog/essentialjsonserializer \cr
#'
#' @docType package
#' @name essentialjsonserializer
NULL
#' EssentialJSONSerializer
#'
#' Provides functions for (de)serialization of R-specific objects to generic JSON datatype objects.
#'
#' Following functions are provided:
#'
#' \itemize{
#' \item RToList: Convert complex R structures to a generic datatype list structure.
#' \item RToJSON: Convert complex R structures to a generic datatype JSON structure.
#' \item RFromList: Convert generic datatype list structure to complex R-specific objects.
#' \item RFromJSON: Convert generic datatype JSON structure to complex R-specific objects.
#' }
#'
#' @examples
#
#' essential.serializer <- EssentialJSONSerializer$new()
#' list.input <- list(c(1, 2, 3, 4, 5),
#' c(6, 7, 8), list(c(1, 2, 3)))
#'
#' list.list <- essential.serializer$RToList(list.input)
#' list.json <- essential.serializer$RToJSON(list.input)
#' list.list.revert <- essential.serializer$RFromList(list.list)
#' list.json.revert <- essential.serializer$RFromJSON(list.json)
#'
#' print(list.input)
#' print(list.json)
#'
#' # datetime
#' datetime.input <- as.POSIXct(Sys.time(), tz="UTC")
#'
#' datetime.list <- essential.serializer$RToList(datetime.input)
#' datetime.json <- essential.serializer$RToJSON(datetime.input)
#' datetime.list.revert <- essential.serializer$RFromList(datetime.list)
#' datetime.json.revert <- essential.serializer$RFromJSON(datetime.json)
#'
#' print(datetime.input)
#' print(datetime.json)
#'
#' # datetime nested
#' datetime.nested.input <- list(input.1 = datetime.input,
#' input.2 = as.POSIXct(Sys.Date(), tz="UTC"))
#'
#' datetime.nested.list <- essential.serializer$RToList(datetime.nested.input)
#' datetime.nested.json <- essential.serializer$RToJSON(datetime.nested.input)
#' datetime.nested.list.revert <- essential.serializer$RFromList(datetime.nested.list)
#' datetime.nested.json.revert <- essential.serializer$RFromJSON(datetime.nested.json)
#'
#' print(datetime.nested.input)
#' print(datetime.nested.json)
#'
#' # timeseries
#' timeseries.input <- xts::xts(matrix(c(5, 7, 9), ncol=1),
#' order.by = c(as.POSIXct("2017-01-01", tz="UTC"),
#' as.POSIXct("2017-01-02", tz="UTC"),
#' as.POSIXct("2017-01-03", tz="UTC")))
#' names(timeseries.input) <- "name_series"
#'
#' timeseries.list <- essential.serializer$RToList(timeseries.input)
#' timeseries.json <- essential.serializer$RToJSON(timeseries.input)
#' timeseries.list.revert <- essential.serializer$RFromList(timeseries.list)
#' timeseries.json.revert <- essential.serializer$RFromJSON(timeseries.json)
#'
#' print(timeseries.input)
#' print(timeseries.json)
#'
#'
#' # timeseries_n_dim
#' timeseriesndim.input <- xts:::xts(matrix(c(5, 7, 9, 12, 15, 19), ncol=2),
#' order.by = c(as.POSIXct("2017-01-01", tz="UTC"),
#' as.POSIXct("2017-01-02", tz="UTC"),
#' as.POSIXct("2017-01-03", tz="UTC")))
#' names(timeseriesndim.input) <- c("name_series_1", "name_series_2")
#'
#' timeseriesndim.list <- essential.serializer$RToList(timeseriesndim.input)
#' timeseriesndim.json <- essential.serializer$RToJSON(timeseriesndim.input)
#' timeseriesndim.list.revert <- essential.serializer$RFromList(timeseriesndim.list)
#' timeseriesndim.json.revert <- essential.serializer$RFromJSON(timeseriesndim.json)
#'
#' print(timeseriesndim.input)
#' print(timeseriesndim.json)
#'
#'
#' # nested timeseries
#' timeseries.nested.input <- list(ts1 = timeseries.input,
#' ts2 = timeseriesndim.input)
#'
#' timeseries.nested.list <- essential.serializer$RToList(timeseries.nested.input)
#' timeseries.nested.json <- essential.serializer$RToJSON(timeseries.nested.input)
#' timeseries.nested.list.revert <- essential.serializer$RFromList(timeseries.nested.list)
#' timeseries.nested.json.revert <- essential.serializer$RFromJSON(timeseries.nested.json)
#'
#' print(timeseries.nested.input)
#' print(timeseries.nested.json)
#'
#' # series
#' series.input <- data.frame(c(1, 2, 3, 4),
#' row.names = c("one", "two", "three", "four"))
#' names(series.input) <- "only_column"
#'
#' series.list <- essential.serializer$RToList(series.input)
#' series.json <- essential.serializer$RToJSON(series.input)
#' series.list.revert <- essential.serializer$RFromList(series.list)
#' series.json.revert <- essential.serializer$RFromJSON(series.json)
#'
#' print(series.input)
#' print(series.json)
#'
#'
#' # dataframe
#' dataframe.input <- data.frame(c(1, 2, 3, 4), c(4, 5, 6, 7),
#' row.names = c("one", "two", "three", "four"))
#' names(dataframe.input) <- c("first_col", "second_col")
#'
#' dataframe.list <- essential.serializer$RToList(dataframe.input)
#' dataframe.json <- essential.serializer$RToJSON(dataframe.input)
#' dataframe.list.revert <- essential.serializer$RFromList(dataframe.list)
#' dataframe.json.revert <- essential.serializer$RFromJSON(dataframe.json)
#'
#' print(dataframe.input)
#' print(dataframe.json)
#' print(dataframe.json.revert)
#'
#' # matrix
#' matrix.input <- matrix(c(1, 2, 3, 4, 5, 6), ncol = 2, byrow = TRUE)
#'
#' matrix.list <- essential.serializer$RToList(datetime.input)
#' matrix.json <- essential.serializer$RToJSON(datetime.input)
#' matrix.list.revert <- essential.serializer$RFromList(datetime.list)
#' matrix.json.revert <- essential.serializer$RFromJSON(datetime.json)
#'
#' print(matrix.input)
#' print(matrix.json)
#'
#'
#' # Complex input
#' to.convert <- list(list.input = list.input,
#' datetime.input = datetime.input,
#' datetime.nested = datetime.nested.input,
#' timeseries.input = timeseries.input,
#' timeseriesndim.input = timeseriesndim.input,
#' series.input = series.input,
#' dataframe.input = dataframe.input,
#' matrix.input = matrix.input,
#' timeseries.nested = timeseries.nested.input)
#'
#'
#' r.list <- essential.serializer$RToList(to.convert)
#' r.json <- essential.serializer$RToJSON(to.convert)
#' r.list.revert <- essential.serializer$RFromList((r.list))
#' r.json.revert <- essential.serializer$RFromJSON(r.json)
#'
#' #print(to.convert)
#' print(r.json)
#'
#' @import xts
#' @export EssentialJSONSerializer
EssentialJSONSerializer <- setRefClass("EssentialJSONSerializer",
methods = list(
RToList = function(to.convert){
# convert R POSIXct to datetime list
if(.DatetimeTypecheck(to.convert)){
to.convert <- .DatetimeToList(to.convert, "datetime")
}
# convert R xts (1-dim) to timeseries list
else if(.TimeseriesTypecheck(to.convert)){
to.convert <- .TimeseriesToList(to.convert, "timeseries")
}
# Convert R xts (n-dim) to timeseries-n-dim list
else if(.TimeseriesNDimTypecheck(to.convert)){
to.convert <- .TimeseriesNDimToList(to.convert, "timeseries_n_dim")
}
# Convert R dataframe (1-dim) to series list
else if(.SeriesTypecheck(to.convert)){
to.convert <- .SeriesToList(to.convert, "series")
}
# Convert R dataframe (n-dim) to dataframe list
else if(.DataframeTypecheck(to.convert)){
to.convert <- .DataframeToList(to.convert, "dataframe")
}
# Convert R matrix to matrix list
else if(.MatrixTypecheck(to.convert)){
to.convert <- .MatrixToList(to.convert, "matrix")
}
else if(is(to.convert, "list") & length(to.convert) != 0){
for(key in seq(1, length(to.convert))){
if(!is.null(to.convert[[key]])){
to.convert[[key]] <- RToList(to.convert[[key]])
}
}
}
return(to.convert)
},
RToJSON = function(to.convert){
return(as.character(jsonlite::toJSON(RToList(to.convert),
auto_unbox=TRUE, digits = 8, na = "null", null = "null")))
},
RFromList = function(to.convert){
# Convert datetime list to R POSIXct
if(.TypeCheckFromList(to.convert, "datetime")){
to.convert <- .DatetimeFromList(to.convert)
}
# Convert timeseries list to R XTS (1-dim)
else if(.TypeCheckFromList(to.convert, "timeseries")){
to.convert <- .TimeseriesFromList(to.convert)
}
# Convert timeseries_n_dim list to R XTS (n-dim)
else if(.TypeCheckFromList(to.convert, "timeseries_n_dim")){
to.convert <- .TimeseriesNDimFromList(to.convert)
}
# Convert series list to R dataframe (1-dim)
else if(.TypeCheckFromList(to.convert, "series")){
to.convert <- .SeriesFromList(to.convert)
}
# Convert datetime list to R dataframe (n-dim)
else if(.TypeCheckFromList(to.convert, "dataframe")){
to.convert <- .DataframeFromList(to.convert)
}
# Convert datetime list to R Matrix
else if(.TypeCheckFromList(to.convert, "matrix")){
to.convert <- .MatrixFromList(to.convert)
}
# recursion
else if(is(to.convert, "list") & length(to.convert) > 0){
for(key in seq(1, length(to.convert))){
if(!is.null(to.convert[[key]])){
to.convert[[key]] <- RFromList(to.convert[[key]])
}
}
}
return(to.convert)
},
RFromJSON = function(to.convert){
return(RFromList(jsonlite::fromJSON(to.convert, simplifyMatrix = FALSE)))
},
#Check from custom type
.TypeCheckFromList = function(input.values, custom.type.name){
if(is(input.values, "list") & "_type" %in% names(input.values)){
if(input.values$'_type' == custom.type.name){
return(TRUE)
}
}
return(FALSE)
},
##########################
###### datetime ##########
##########################
# check language specific datetime format
.DatetimeTypecheck = function(input.values){
return(is(input.values, "POSIXct"))
},
# Convert lanugage specific datetime object to datetime dict
.DatetimeToList = function(datetime.r, custom.type.name){
datetime.list <- list("_type" = custom.type.name,
"_data" = strftime(datetime.r, format="%Y-%m-%dT%H:%M:%OS"))
return(datetime.list)
},
# convert datetime dict to language specific datetime object
.DatetimeFromList = function(input.values){
return(as.POSIXct(input.values$`_data`, format = "%Y-%m-%dT%H:%M:%OS", tz="UTC"))
},
##########################
###### timeseries ########
##########################
# check language specific timeseries format
.TimeseriesTypecheck = function(input.values){
if(!is(input.values, "xts")){
return(FALSE)
}
if(!(ncol(input.values) == 1)){
return(FALSE)
}
return(TRUE)
},
# Convert language specific timeseries object to timeseries dict
.TimeseriesToList = function(timeseries.r, custom.type.name){
series.name <- names(timeseries.r)
series.timestamps <- strftime(zoo::index(timeseries.r), format="%Y-%m-%dT%H:%M:%OS")
series.values <- as.vector(zoo::coredata(timeseries.r))
timeseries.data <- list("_timestamps" = series.timestamps,
series.name = series.values)
# update column name
names(timeseries.data) <- c("_timestamps", series.name)
timeseries.list <- list("_type" = custom.type.name,
"_data" = timeseries.data)
return(timeseries.list)
},
# Convert timeseries dict to language specific timeseries object
.TimeseriesFromList = function(input.values){
timeseries.list <- input.values$`_data`
# Extract timeindex
xts.index <- as.POSIXct(timeseries.list$`_timestamps`,
format = "%Y-%m-%dT%H:%M:%OS", tz="UTC")
# Extract values
list.names <- names(timeseries.list)
xts.key <- list.names[which(list.names != "_timestamps")]
xts.values <- as.matrix(timeseries.list[[xts.key]], ncol=1)
xts.timeseries <- xts::xts(xts.values, order.by = xts.index)
names(xts.timeseries) <- xts.key
return(xts.timeseries)
},
##################################
###### timeseries_n_dim ##########
##################################
# check language specific timeseries_n_dim format
.TimeseriesNDimTypecheck = function(input.values){
if(!is(input.values, "xts")){
return(FALSE)
}
if(!(ncol(input.values) > 1)){
return(FALSE)
}
return(TRUE)
},
# Convert language specific timeseries object to timeseries dict
.TimeseriesNDimToList = function(timeseries.r, custom.type.name){
series.names <- names(timeseries.r)
series.timestamps <- strftime(zoo::index(timeseries.r), format="%Y-%m-%dT%H:%M:%OS")
series.values <- zoo::coredata(timeseries.r)
timeseries.data <- list("_timestamps" = series.timestamps)
for(i in seq(1, ncol(series.values))){
timeseries.data[[i+1]] <- series.values[,i]
}
# update column name
names(timeseries.data) <- c("_timestamps", series.names)
timeseries.list <- list("_type" = custom.type.name,
"_data" = timeseries.data)
return(timeseries.list)
},
# Convert timeseries dict to language specific timeseries object
.TimeseriesNDimFromList = function(input.values){
timeseries.list <- input.values$`_data`
# Extract timeindex
xts.index <- as.POSIXct(timeseries.list$`_timestamps`,
format = "%Y-%m-%dT%H:%M:%OS", tz="UTC")
# Extract values
xts.timeseries <- xts::xts(, order.by = xts.index)
list.names <- names(timeseries.list)
xts.keys <- list.names[which(list.names != "_timestamps")]
for(key in xts.keys){
xts.values.key <- as.matrix(timeseries.list[[key]], ncol=1)
xts.timeseries <- cbind(xts.timeseries, xts.values.key)
}
names(xts.timeseries) <- xts.keys
return(xts.timeseries)
},
##################################
###### series ####################
##################################
# check language specific series format
.SeriesTypecheck = function(input.values){
if(!is(input.values, "data.frame")){
return(FALSE)
}
if(!(ncol(input.values) == 1)){
return(FALSE)
}
return(TRUE)
},
# Convert language specific timeseries object to timeseries dict
.SeriesToList = function(df.r, custom.type.name = "series"){
# Extract series name
df.name <- names(df.r)
# Extract series values
df.values <- df.r[,1]
# Extract series index
df.index <- rownames(df.r)
# Construct data list
list.data <- list("_index" = df.index,
values = df.values)
names(list.data) <- c("_index", df.name)
# Consstruct series list
df.list <- list("_type" = custom.type.name,
"_data" = list.data)
return(df.list)
},
# Convert series dict to language specific timeseries object
.SeriesFromList = function(input.values){
list.data <- input.values$`_data`
# Extract index
df.index <- list.data$`_index`
# Extract values
list.names <- names(list.data)
df.name <- list.names[which(list.names != "_index")]
df.values <- list.data[[df.name]]
# Construct dataframe
df.series <- data.frame(df.values, row.names = df.index)
names(df.series) <- df.name
return(df.series)
},
##################################
###### dataframe #################
##################################
# check language specific datetime format
.DataframeTypecheck = function(input.values){
if(!is(input.values, "data.frame")){
return(FALSE)
}
if(!(ncol(input.values) > 1)){
return(FALSE)
}
return(TRUE)
},
# Convert language specific dataframe object to dataframe dict
.DataframeToList = function(df.r, custom.type.name = "dataframe"){
# Extract series name
df.names <- names(df.r)
df.index <- rownames(df.r)
# Construct data list
list.data <- list("_index" = df.index)
# Construct
for(i in seq(1, ncol(df.r))){
list.data[[i+1]] <- df.r[,i]
}
names(list.data) <- c("_index", df.names)
# Consstruct series list
df.list <- list("_type" = custom.type.name,
"_data" = list.data)
return(df.list)
},
# Convert dataframe dict to language specific dataframe object
.DataframeFromList = function(input.values){
list.data <- input.values$`_data`
# Extract index
df.index <- list.data$`_index`
# empty dataframe
if(length(df.index) == 0)
return(data.frame())
# Extract values
list.names <- names(list.data)
df.names <- list.names[which(list.names != "_index")]
# Construct dataframe
df.series <- data.frame(list.data[[df.names[1]]], row.names = df.index)
for(i in seq(2, length(df.names))){
df.series <- cbind(df.series, list.data[[df.names[i]]])
}
names(df.series) <- df.names
return(df.series)
},
##################################
###### matrix ####################
##################################
# check language specific datetime format
.MatrixTypecheck = function(input.values){
return(is(input.values, "matrix"))
},
.MatrixToList = function(matrix.r, custom.type.name = "matrix"){
list.data <- list(matrix.r[1, ])
for(i in seq(2, nrow(matrix.r))){
list.data[[i]] <- matrix.r[i,]
}
matrix.list <- list("_type" = custom.type.name,
"_data" = list.data)
return(matrix.list)
},
.MatrixFromList = function(input.values){
return(do.call("rbind", input.values$`_data`))
}
))
VermeirJellen/essentialjsonserializer_r documentation built on May 16, 2019, 4:24 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Sys.setenv(tz="UTC")
#' JSON serialization of complex R-objects \cr
#'
#' The intention of the package fuctionality is to (de)serialize language specific complex objects towards generic
#' JSON-datatype objects. The JSON conversion allows for flexible cross language communication when the package is implemented
#' in different languages. This package provides an R implementation of the (de)serialization functionality.
#'
#'
#' Currently, 6 generic complex JSON-datatypes are provided. Corresponding R-mappings are defined as follows:
#'
#' \itemize{
#' \item JSON datetime - maps to POSIXct objects
#' \item JSON timeseries - maps to 1-dimensional xts objects
#' \item JSON timeseries_n_dim - maps to n-dimensional xts objects (n > 1)
#' \item JSON series: - maps to 1-dimensional dataframe objects
#' \item JSON dataframe: - maps to n-dimensional dataframe objects (n > 1)
#' \item JSON matrix: - maps to matrix objects
#' }
#'
#' The package provides the user with the EssentialJSONSerializer reference class. Following functions
#' are provided:
#' \itemize{
#' \item RToList: Convert complex R structures to a generic datatype list structure.
#' \item RToJSON: Convert complex R structures to a generic datatype JSON structure.
#' \item RFromList: Convert generic datatype list structure to complex R-specific objects.
#' \item RFromJSON: Convert generic datatype JSON structure to complex R-specific objects.
#' }
#'
#' Examples: \cr
#' - https://github.com/VermeirJellen/essentialjsonserializer_r \cr
#' - http://EssentialDataScience.com/blog/essentialjsonserializer \cr
#'
#' @docType package
#' @name essentialjsonserializer
NULL
#' EssentialJSONSerializer
#'
#' Provides functions for (de)serialization of R-specific objects to generic JSON datatype objects.
#'
#' Following functions are provided:
#'
#' \itemize{
#' \item RToList: Convert complex R structures to a generic datatype list structure.
#' \item RToJSON: Convert complex R structures to a generic datatype JSON structure.
#' \item RFromList: Convert generic datatype list structure to complex R-specific objects.
#' \item RFromJSON: Convert generic datatype JSON structure to complex R-specific objects.
#' }
#'
#' @examples
#
#' essential.serializer <- EssentialJSONSerializer$new()
#' list.input <- list(c(1, 2, 3, 4, 5),
#' c(6, 7, 8), list(c(1, 2, 3)))
#'
#' list.list <- essential.serializer$RToList(list.input)
#' list.json <- essential.serializer$RToJSON(list.input)
#' list.list.revert <- essential.serializer$RFromList(list.list)
#' list.json.revert <- essential.serializer$RFromJSON(list.json)
#'
#' print(list.input)
#' print(list.json)
#'
#' # datetime
#' datetime.input <- as.POSIXct(Sys.time(), tz="UTC")
#'
#' datetime.list <- essential.serializer$RToList(datetime.input)
#' datetime.json <- essential.serializer$RToJSON(datetime.input)
#' datetime.list.revert <- essential.serializer$RFromList(datetime.list)
#' datetime.json.revert <- essential.serializer$RFromJSON(datetime.json)
#'
#' print(datetime.input)
#' print(datetime.json)
#'
#' # datetime nested
#' datetime.nested.input <- list(input.1 = datetime.input,
#' input.2 = as.POSIXct(Sys.Date(), tz="UTC"))
#'
#' datetime.nested.list <- essential.serializer$RToList(datetime.nested.input)
#' datetime.nested.json <- essential.serializer$RToJSON(datetime.nested.input)
#' datetime.nested.list.revert <- essential.serializer$RFromList(datetime.nested.list)
#' datetime.nested.json.revert <- essential.serializer$RFromJSON(datetime.nested.json)
#'
#' print(datetime.nested.input)
#' print(datetime.nested.json)
#'
#' # timeseries
#' timeseries.input <- xts::xts(matrix(c(5, 7, 9), ncol=1),
#' order.by = c(as.POSIXct("2017-01-01", tz="UTC"),
#' as.POSIXct("2017-01-02", tz="UTC"),
#' as.POSIXct("2017-01-03", tz="UTC")))
#' names(timeseries.input) <- "name_series"
#'
#' timeseries.list <- essential.serializer$RToList(timeseries.input)
#' timeseries.json <- essential.serializer$RToJSON(timeseries.input)
#' timeseries.list.revert <- essential.serializer$RFromList(timeseries.list)
#' timeseries.json.revert <- essential.serializer$RFromJSON(timeseries.json)
#'
#' print(timeseries.input)
#' print(timeseries.json)
#'
#'
#' # timeseries_n_dim
#' timeseriesndim.input <- xts:::xts(matrix(c(5, 7, 9, 12, 15, 19), ncol=2),
#' order.by = c(as.POSIXct("2017-01-01", tz="UTC"),
#' as.POSIXct("2017-01-02", tz="UTC"),
#' as.POSIXct("2017-01-03", tz="UTC")))
#' names(timeseriesndim.input) <- c("name_series_1", "name_series_2")
#'
#' timeseriesndim.list <- essential.serializer$RToList(timeseriesndim.input)
#' timeseriesndim.json <- essential.serializer$RToJSON(timeseriesndim.input)
#' timeseriesndim.list.revert <- essential.serializer$RFromList(timeseriesndim.list)
#' timeseriesndim.json.revert <- essential.serializer$RFromJSON(timeseriesndim.json)
#'
#' print(timeseriesndim.input)
#' print(timeseriesndim.json)
#'
#'
#' # nested timeseries
#' timeseries.nested.input <- list(ts1 = timeseries.input,
#' ts2 = timeseriesndim.input)
#'
#' timeseries.nested.list <- essential.serializer$RToList(timeseries.nested.input)
#' timeseries.nested.json <- essential.serializer$RToJSON(timeseries.nested.input)
#' timeseries.nested.list.revert <- essential.serializer$RFromList(timeseries.nested.list)
#' timeseries.nested.json.revert <- essential.serializer$RFromJSON(timeseries.nested.json)
#'
#' print(timeseries.nested.input)
#' print(timeseries.nested.json)
#'
#' # series
#' series.input <- data.frame(c(1, 2, 3, 4),
#' row.names = c("one", "two", "three", "four"))
#' names(series.input) <- "only_column"
#'
#' series.list <- essential.serializer$RToList(series.input)
#' series.json <- essential.serializer$RToJSON(series.input)
#' series.list.revert <- essential.serializer$RFromList(series.list)
#' series.json.revert <- essential.serializer$RFromJSON(series.json)
#'
#' print(series.input)
#' print(series.json)
#'
#'
#' # dataframe
#' dataframe.input <- data.frame(c(1, 2, 3, 4), c(4, 5, 6, 7),
#' row.names = c("one", "two", "three", "four"))
#' names(dataframe.input) <- c("first_col", "second_col")
#'
#' dataframe.list <- essential.serializer$RToList(dataframe.input)
#' dataframe.json <- essential.serializer$RToJSON(dataframe.input)
#' dataframe.list.revert <- essential.serializer$RFromList(dataframe.list)
#' dataframe.json.revert <- essential.serializer$RFromJSON(dataframe.json)
#'
#' print(dataframe.input)
#' print(dataframe.json)
#' print(dataframe.json.revert)
#'
#' # matrix
#' matrix.input <- matrix(c(1, 2, 3, 4, 5, 6), ncol = 2, byrow = TRUE)
#'
#' matrix.list <- essential.serializer$RToList(datetime.input)
#' matrix.json <- essential.serializer$RToJSON(datetime.input)
#' matrix.list.revert <- essential.serializer$RFromList(datetime.list)
#' matrix.json.revert <- essential.serializer$RFromJSON(datetime.json)
#'
#' print(matrix.input)
#' print(matrix.json)
#'
#'
#' # Complex input
#' to.convert <- list(list.input = list.input,
#' datetime.input = datetime.input,
#' datetime.nested = datetime.nested.input,
#' timeseries.input = timeseries.input,
#' timeseriesndim.input = timeseriesndim.input,
#' series.input = series.input,
#' dataframe.input = dataframe.input,
#' matrix.input = matrix.input,
#' timeseries.nested = timeseries.nested.input)
#'
#'
#' r.list <- essential.serializer$RToList(to.convert)
#' r.json <- essential.serializer$RToJSON(to.convert)
#' r.list.revert <- essential.serializer$RFromList((r.list))
#' r.json.revert <- essential.serializer$RFromJSON(r.json)
#'
#' #print(to.convert)
#' print(r.json)
#'
#' @import xts
#' @export EssentialJSONSerializer
EssentialJSONSerializer <- setRefClass("EssentialJSONSerializer",
methods = list(
RToList = function(to.convert){
# convert R POSIXct to datetime list
if(.DatetimeTypecheck(to.convert)){
to.convert <- .DatetimeToList(to.convert, "datetime")
}
# convert R xts (1-dim) to timeseries list
else if(.TimeseriesTypecheck(to.convert)){
to.convert <- .TimeseriesToList(to.convert, "timeseries")
}
# Convert R xts (n-dim) to timeseries-n-dim list
else if(.TimeseriesNDimTypecheck(to.convert)){
to.convert <- .TimeseriesNDimToList(to.convert, "timeseries_n_dim")
}
# Convert R dataframe (1-dim) to series list
else if(.SeriesTypecheck(to.convert)){
to.convert <- .SeriesToList(to.convert, "series")
}
# Convert R dataframe (n-dim) to dataframe list
else if(.DataframeTypecheck(to.convert)){
to.convert <- .DataframeToList(to.convert, "dataframe")
}
# Convert R matrix to matrix list
else if(.MatrixTypecheck(to.convert)){
to.convert <- .MatrixToList(to.convert, "matrix")
}
else if(is(to.convert, "list") & length(to.convert) != 0){
for(key in seq(1, length(to.convert))){
if(!is.null(to.convert[[key]])){
to.convert[[key]] <- RToList(to.convert[[key]])
}
}
}
return(to.convert)
},
RToJSON = function(to.convert){
return(as.character(jsonlite::toJSON(RToList(to.convert),
auto_unbox=TRUE, digits = 8, na = "null", null = "null")))
},
RFromList = function(to.convert){
# Convert datetime list to R POSIXct
if(.TypeCheckFromList(to.convert, "datetime")){
to.convert <- .DatetimeFromList(to.convert)
}
# Convert timeseries list to R XTS (1-dim)
else if(.TypeCheckFromList(to.convert, "timeseries")){
to.convert <- .TimeseriesFromList(to.convert)
}
# Convert timeseries_n_dim list to R XTS (n-dim)
else if(.TypeCheckFromList(to.convert, "timeseries_n_dim")){
to.convert <- .TimeseriesNDimFromList(to.convert)
}
# Convert series list to R dataframe (1-dim)
else if(.TypeCheckFromList(to.convert, "series")){
to.convert <- .SeriesFromList(to.convert)
}
# Convert datetime list to R dataframe (n-dim)
else if(.TypeCheckFromList(to.convert, "dataframe")){
to.convert <- .DataframeFromList(to.convert)
}
# Convert datetime list to R Matrix
else if(.TypeCheckFromList(to.convert, "matrix")){
to.convert <- .MatrixFromList(to.convert)
}
# recursion
else if(is(to.convert, "list") & length(to.convert) > 0){
for(key in seq(1, length(to.convert))){
if(!is.null(to.convert[[key]])){
to.convert[[key]] <- RFromList(to.convert[[key]])
}
}
}
return(to.convert)
},
RFromJSON = function(to.convert){
return(RFromList(jsonlite::fromJSON(to.convert, simplifyMatrix = FALSE)))
},
#Check from custom type
.TypeCheckFromList = function(input.values, custom.type.name){
if(is(input.values, "list") & "_type" %in% names(input.values)){
if(input.values$'_type' == custom.type.name){
return(TRUE)
}
}
return(FALSE)
},
##########################
###### datetime ##########
##########################
# check language specific datetime format
.DatetimeTypecheck = function(input.values){
return(is(input.values, "POSIXct"))
},
# Convert lanugage specific datetime object to datetime dict
.DatetimeToList = function(datetime.r, custom.type.name){
datetime.list <- list("_type" = custom.type.name,
"_data" = strftime(datetime.r, format="%Y-%m-%dT%H:%M:%OS"))
return(datetime.list)
},
# convert datetime dict to language specific datetime object
.DatetimeFromList = function(input.values){
return(as.POSIXct(input.values$`_data`, format = "%Y-%m-%dT%H:%M:%OS", tz="UTC"))
},
##########################
###### timeseries ########
##########################
# check language specific timeseries format
.TimeseriesTypecheck = function(input.values){
if(!is(input.values, "xts")){
return(FALSE)
}
if(!(ncol(input.values) == 1)){
return(FALSE)
}
return(TRUE)
},
# Convert language specific timeseries object to timeseries dict
.TimeseriesToList = function(timeseries.r, custom.type.name){
series.name <- names(timeseries.r)
series.timestamps <- strftime(zoo::index(timeseries.r), format="%Y-%m-%dT%H:%M:%OS")
series.values <- as.vector(zoo::coredata(timeseries.r))
timeseries.data <- list("_timestamps" = series.timestamps,
series.name = series.values)
# update column name
names(timeseries.data) <- c("_timestamps", series.name)
timeseries.list <- list("_type" = custom.type.name,
"_data" = timeseries.data)
return(timeseries.list)
},
# Convert timeseries dict to language specific timeseries object
.TimeseriesFromList = function(input.values){
timeseries.list <- input.values$`_data`
# Extract timeindex
xts.index <- as.POSIXct(timeseries.list$`_timestamps`,
format = "%Y-%m-%dT%H:%M:%OS", tz="UTC")
# Extract values
list.names <- names(timeseries.list)
xts.key <- list.names[which(list.names != "_timestamps")]
xts.values <- as.matrix(timeseries.list[[xts.key]], ncol=1)
xts.timeseries <- xts::xts(xts.values, order.by = xts.index)
names(xts.timeseries) <- xts.key
return(xts.timeseries)
},
##################################
###### timeseries_n_dim ##########
##################################
# check language specific timeseries_n_dim format
.TimeseriesNDimTypecheck = function(input.values){
if(!is(input.values, "xts")){
return(FALSE)
}
if(!(ncol(input.values) > 1)){
return(FALSE)
}
return(TRUE)
},
# Convert language specific timeseries object to timeseries dict
.TimeseriesNDimToList = function(timeseries.r, custom.type.name){
series.names <- names(timeseries.r)
series.timestamps <- strftime(zoo::index(timeseries.r), format="%Y-%m-%dT%H:%M:%OS")
series.values <- zoo::coredata(timeseries.r)
timeseries.data <- list("_timestamps" = series.timestamps)
for(i in seq(1, ncol(series.values))){
timeseries.data[[i+1]] <- series.values[,i]
}
# update column name
names(timeseries.data) <- c("_timestamps", series.names)
timeseries.list <- list("_type" = custom.type.name,
"_data" = timeseries.data)
return(timeseries.list)
},
# Convert timeseries dict to language specific timeseries object
.TimeseriesNDimFromList = function(input.values){
timeseries.list <- input.values$`_data`
# Extract timeindex
xts.index <- as.POSIXct(timeseries.list$`_timestamps`,
format = "%Y-%m-%dT%H:%M:%OS", tz="UTC")
# Extract values
xts.timeseries <- xts::xts(, order.by = xts.index)
list.names <- names(timeseries.list)
xts.keys <- list.names[which(list.names != "_timestamps")]
for(key in xts.keys){
xts.values.key <- as.matrix(timeseries.list[[key]], ncol=1)
xts.timeseries <- cbind(xts.timeseries, xts.values.key)
}
names(xts.timeseries) <- xts.keys
return(xts.timeseries)
},
##################################
###### series ####################
##################################
# check language specific series format
.SeriesTypecheck = function(input.values){
if(!is(input.values, "data.frame")){
return(FALSE)
}
if(!(ncol(input.values) == 1)){
return(FALSE)
}
return(TRUE)
},
# Convert language specific timeseries object to timeseries dict
.SeriesToList = function(df.r, custom.type.name = "series"){
# Extract series name
df.name <- names(df.r)
# Extract series values
df.values <- df.r[,1]
# Extract series index
df.index <- rownames(df.r)
# Construct data list
list.data <- list("_index" = df.index,
values = df.values)
names(list.data) <- c("_index", df.name)
# Consstruct series list
df.list <- list("_type" = custom.type.name,
"_data" = list.data)
return(df.list)
},
# Convert series dict to language specific timeseries object
.SeriesFromList = function(input.values){
list.data <- input.values$`_data`
# Extract index
df.index <- list.data$`_index`
# Extract values
list.names <- names(list.data)
df.name <- list.names[which(list.names != "_index")]
df.values <- list.data[[df.name]]
# Construct dataframe
df.series <- data.frame(df.values, row.names = df.index)
names(df.series) <- df.name
return(df.series)
},
##################################
###### dataframe #################
##################################
# check language specific datetime format
.DataframeTypecheck = function(input.values){
if(!is(input.values, "data.frame")){
return(FALSE)
}
if(!(ncol(input.values) > 1)){
return(FALSE)
}
return(TRUE)
},
# Convert language specific dataframe object to dataframe dict
.DataframeToList = function(df.r, custom.type.name = "dataframe"){
# Extract series name
df.names <- names(df.r)
df.index <- rownames(df.r)
# Construct data list
list.data <- list("_index" = df.index)
# Construct
for(i in seq(1, ncol(df.r))){
list.data[[i+1]] <- df.r[,i]
}
names(list.data) <- c("_index", df.names)
# Consstruct series list
df.list <- list("_type" = custom.type.name,
"_data" = list.data)
return(df.list)
},
# Convert dataframe dict to language specific dataframe object
.DataframeFromList = function(input.values){
list.data <- input.values$`_data`
# Extract index
df.index <- list.data$`_index`
# empty dataframe
if(length(df.index) == 0)
return(data.frame())
# Extract values
list.names <- names(list.data)
df.names <- list.names[which(list.names != "_index")]
# Construct dataframe
df.series <- data.frame(list.data[[df.names[1]]], row.names = df.index)
for(i in seq(2, length(df.names))){
df.series <- cbind(df.series, list.data[[df.names[i]]])
}
names(df.series) <- df.names
return(df.series)
},
##################################
###### matrix ####################
##################################
# check language specific datetime format
.MatrixTypecheck = function(input.values){
return(is(input.values, "matrix"))
},
.MatrixToList = function(matrix.r, custom.type.name = "matrix"){
list.data <- list(matrix.r[1, ])
for(i in seq(2, nrow(matrix.r))){
list.data[[i]] <- matrix.r[i,]
}
matrix.list <- list("_type" = custom.type.name,
"_data" = list.data)
return(matrix.list)
},
.MatrixFromList = function(input.values){
return(do.call("rbind", input.values$`_data`))
}
))
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.